Oral squamous cell carcinomas (including those of the mouth and oropharynx) comprise more than 90% of all cancers affecting these tissues. Of all the major cancers, oral cancer has one of the worst five-year survival rates of 62.7%. Historically, the death rate associated with this cancer is particularly high because it is discovered late (stage III or IV), often when it has spread to the lymph nodes or deeper structures. Late stage cancers have high mortality, are expensive to treat and have high morbidity. The current gold standard for oral cancer diagnosis is a physical and visual examination followed by biopsy. Unfortunately, due to delays in diagnosis, most cases are diagnosed in late stage with a survival rate around 20-40% even after aggressive treatment regimens including combinations of radiation, surgery and chemotherapy. However, if diagnosed early (stage I/II), the five-year survival-rate can be as high as 80-90%, representing a significant healthcare cost savings. Thus, there is a pressing need for an improved oral cancer early detection test which will save lives, improve clinical outcome and reduce treatment costs. We have developed a unique, easy to use, oral cancer specific, rapid lateral flow point-of-care (POC) test based on the tumor initiating marker CD44 and total protein that accurately assesses patient's risk for oral cancer. The OncAlert lateral flow test strip identifies those individuals a most risk of harboring early or developing oral cancer from the 85 million tobacco and alcohol users. However, often these early lesions can be invisible to the naked eye, so the test strip alone will not indicate where the abnormal cells are in cases of very early disease. To answer this pressing medical need, in this application, we will develop an imaging device that will involve a simple rinse and spit solution with an antiCD44 probe bound to non-absorbable microspheres that will bind to cells expressing abnormally high levels of CD44. The bound probe can then be visualized by shining a light of specific wavelength so that very early lesions can be identified. If successful, this is expected to enable early detection and intervention of orl cancer even be- fore visual symptoms are revealed, thus addressing a crucial clinically unmet need. We will carry out the fol- lowing specific aims: in Specific Aim #1 we will test an in-vitro model for detecting CD44 elevated cells using fluorescent probes on membranes. We will determine whether this potential imaging agent recognizes CD44 on the surface of cells. To do this we will culture well-characterized cell lines that have been engineered to overexpress or underexpress CD44. In Specific Aim #2 we will test the system in tumor explants from humans. We will determine how well the conjugate identifies carcinogenesis in humans. We will use human up- per aero-digestive tract specimens ex vivo, immediately following surgical resection. We will analyze 10 oral cancer tissues with surrounding normal margin and 10 patient tissues that were excised for benign conditions (chronic tonsillitis) using the anti-CD44 fluorescently labeled beads.